this page left intentionally blank
RESEARCH METHODS
in the SOCIAL SCIENCES
this page left intentionally blank
Eighth Edition
RESEARCH METHODS
in the SOCIAL SCIENCES
Chava Frankfort-Nachmias
University of Wisconsin, Milwaukee
David Nachmias
The Interdisciplinary Center, Israel
Jack DeWaard
University of Minnesota
Worth Publishers
A Macmillan Education Company
To our daughters, Anat and Talia
—Chava and David
Senior Vice President, Editorial and Production: Catherine Woods
Director, Print and Digital Production Development: Tracey Kuehn
Publisher: Rachel Losh
Associate Publisher: Jessica Bayne
Senior Associate Editor: Sarah Berger
Senior Project Editor: Vivien Weiss
Development Editor: Thomas Finn
Editorial Assistant: Kimberly Morgan
Executive Marketing Manager: Katherine Nurre
Marketing Assistant: Allison Greco
Design Manager: Blake Logan
Text Designer: Kevin Kall
Art Manager: Matthew McAdams
Production Manager: Barbara Seixas
Media and Supplements Editor: Anthony Casciano
Composition: TSI evolve
Printing and Binding: RR Donnelley
Cover Image: RTimages/Shutterstock
Library of Congress Control Number: 2014948078
ISBN-13: 978-1-4292-3300-2
ISBN-10: 1-4292-3300-1
© 2015, 2008, 1999, 1996 by Worth Publishers. All rights reserved.
Printed in the United States of America
First printing
Worth Publishers
41 Madison Avenue
New York, NY 10010
www.worthpublishers.com
Contents in Brief
PREFACE xi
Part I
Foundations of Empirical Research 1
1 The Scientific Approach 3
2 Conceptual Foundations of Research 23
3 Elements of Research 45
4 Ethics in Social Sciences Research 63
Part II
Design and Structure of Research 79
5 Research Designs: Experiments 81
6 Research Designs: Cross-Sectional and Quasi-Experimental
Designs 103
7 Measurement 121
8 Sampling and Sample Designs 143
Part III Data Collection 167
9 Observational Methods 169
10 Survey Research 187
11 Questionnaire Construction 213
12 Qualitative Research 241
13 Secondary Data and Content Analysis 261
Part IV Data Processing and Analysis 285
14 Data Preparation and Analysis 287
15 The Univariate Distribution 303
16 Bivariate Analysis 331
17 Multivariate Analysis 359
18 Index Construction and Scaling Methods 383
19 Inferences 403
Appendices
A Writing Research Reports 425
B ∑: The Summation Sign 437
C Random Digits 439
D Areas Under the Normal Curve 443
E Distribution of t 444
F Critical Values of F 445
G Distribution of x 2 450
Glossary 453
Author Index AI-1
Subject Index SI-1
v
this page left intentionally blank
Contents
PREFACE xi
Part I
Foundations of Empirical Research 1
1
The Scientific Approach 3
What Is Science? 4
Approaches to Knowledge 4
Basic Assumptions of Science 6
Aims of the Social Sciences 8
The Roles of Methodology 12
Scientific Revolutions 14
The Research Process 17
The Plan of This Book 18
2
Conceptual Foundations of Research 23
Concepts 24
Definitions 26
Theory: Functions and Types 29
Models 35
Theory, Models, and Empirical Research 38
3
Elements of Research 45
Research Problems 46
Units of Analysis 47
Variables 49
Relations 52
Hypotheses 55
Research Questions and Hypotheses: An Example 56
4
Ethics in Social Sciences Research 63
Why Research Ethics? 64
Balancing Costs and Benefits 67
Informed Consent 67
Privacy 70
Professional Codes of Ethics 72
vii
viii
CONTENTS
Part II
Design and Structure of Research 79
5
Research Designs: Experiments 81
Classic Experimental Research Design 82
Causal Inferences 85
Components of a Research Design 86
Design Types 94
6
Research Designs: Cross-Sectional and
Quasi-Experimental Designs 103
Types of Relations and Designs 104
Cross-Sectional Designs 105
Quasi-Experimental Designs 106
Combined Designs 114
Preexperimental Designs 115
A Comparison of Designs 116
7
Measurement 121
The Process of Measurement 122
Levels of Measurement 125
Data Transformation 129
Measurement Error 130
Validity 131
Reliability 135
8
Sampling and Sample Designs 143
Aims of Sampling 144
The Population 145
Sample Designs 147
Sample Size 156
Nonsampling Errors 161
Part III
Data Collection 167
9
Observational Methods 169
Triangulation 170
Roles of Observation 171
Types of Behavior 172
Timing and Recording 175
Inference in the Course of Observation 176
Controlled Observation 176
CONTENTS
10
Survey Research 187
Mail Questionnaire 188
Personal Interview 194
Telephone Interview 202
Online Surveys, Live Feeds, and Big Data 204
Comparing the Four Survey Methods 206
Conclusion 207
11
Questionnaire Construction 213
The Question 214
Content of Questions 214
Types of Questions 217
Question Format 220
Sequence of Questions 222
Avoiding Bias: Pitfalls in Questionnaire Construction 225
Introductions 227
Instructions 229
Constructing a Questionnaire: A Case Study 230
12
Qualitative Research 241
Field Research 243
Participant Observation 244
Ethnography 247
The Practice of Qualitative Research 248
Blue-Collar Community: Qualitative Research in Practice 254
Ethical and Political Issues of Qualitative Research 256
13
Secondary Data and Content Analysis 261
Why Secondary Data Analysis? 262
Searching and Sourcing Secondary Data 265
Content Analysis 273
Part IV
Data Processing and Analysis 285
14
Data Preparation and Analysis 287
Coding Schemes 288
Codebook Construction 293
Computing in Social Sciences Research 297
15
The Univariate Distribution 303
The Role of Statistics 304
ix
x
CONTENTS
Frequency Distributions 305
Using Graphs to Describe Distributions 308
Measures of Central Tendency 311
Basic Measures of Dispersion 317
Measures of Dispersion Based on the Mean 319
Types of Frequency Distributions 322
16
Bivariate Analysis 331
The Concept of Relationship 332
Nominal Measures of Relationship 338
Ordinal Measures of Relationship 341
Interval Measures of Relationship 347
17
Multivariate Analysis 359
Analyzing Multiple Variables 361
Examining Relationships Among Variables 363
Statistical Procedures 366
18
Index Construction and Scaling Methods 383
Index Construction 385
Scaling Methods 391
19
Inferences 403
The Strategy of Testing Hypotheses 405
Null and Research Hypotheses 406
Sampling Distribution 406
Level of Significance and Region of Rejection 408
Parametric and Nonparametric Tests of Significance 412
Appendices
A Writing Research Reports 425
B ∑: The Summation Sign 437
C Random Digits 439
D Areas Under the Normal Curve 443
E Distribution of t 444
F Critical Values of F 445
G Distribution of x 2 450
Glossary 453
Author Index AI-1
Subject Index SI-1
Preface
The goal of the eighth edition of Research Methods in the Social Sciences, as in the previous
editions, is to offer a comprehensive, systematic presentation of the scientific approach
within the context of the social sciences. We emphasize the relationship among theory,
research, and practice and integrate research activities in an orderly framework so that
the reader can more easily learn about the nature of social sciences research.
In our view, social sciences research is a cyclical, self-correcting process consisting
of seven major interrelated stages: definition of the research problem, statement of the
hypothesis, research design, measurement, data collection, data analysis, and generalization. Each of these stages is interrelated with theory. The text leads the reader through
each stage of this process.
The New Edition
The new edition continues to blend a broad range of classic social sciences research
studies with up-to-date examples of contemporary social sciences issues. The additions
and changes reflect developments in the field since publication of the previous edition.
Major updates and revisions to the eighth edition include the following:
O New Part Openers preview the concepts covered in each of the four major
sections of the book.
O Lengthier chapters from the previous edition have been streamlined to
highlight central concepts.
O Current and interdisciplinary examples draw from a range of social sciences
disciplines, including anthropology, economics, geography, history, political
science, psychology, public policy, and sociology, among others.
O Expanded coverage of research methods in the digital age, including the
use of the Internet and various computer software packages for retrieving,
cleaning, coding, and analyzing “big data.”
O Expanded coverage of qualitative research methods, most especially of
participant observation and ethnography.
The following updates and revisions have likewise been implemented in each chapter:
O New chapter introductions illustrate the central themes in each chapter,
drawing on current topics and research studies, including the use of Twitter
feeds in social sciences research, the 2011 revolutions in Egypt and Tunisia
(i.e., the “Arab Spring”), and climate change.
O Updated research examples are provided throughout each chapter.
O Chapter examples using data from the General Social Survey (GSS) have
been updated to reflect data from the 2012 round of the GSS.
O Expanded “Study Questions” are provided at the end of each chapter.
O New end-of-chapter exercises on “Reading and Writing Research Reports”
help students understand and apply chapter concepts to the practice and
process of social sciences research.
xi
xii
PREFACE
The Plan of This Book
The book’s organization progresses logically from the conceptual and theoretical building blocks of the research process to data analysis and computer applications, offering
students a comprehensive and systematic foundation for comprehending the breadth
and depth of social sciences research. The book’s self-contained yet integrated chapters
promote flexibility in structuring courses depending upon the individual instructor’s
needs and interests. The text adapts easily to two kinds of courses: a basic research
methods course or one that covers research methods and statistics sequentially.
Chapter 1 examines the foundations of knowledge, the objectives of scientific research, and the basic assumptions of the scientific approach. Chapters 2 and 3 discuss
the basic issues of empirical research and the relationship between theory and research.
They cover the topics of concept formation, the roles and types of theories, models,
variables, and hypothesis construction. Chapter 4 focuses on ethical concerns in social
sciences research and considers ways to ensure the rights and the welfare of research
participants, including the right to privacy.
Chapters 5 and 6 present the research design stage. A research design is a strategy
that guides investigators; it is a logical model for inferring causal relations. Experimental
designs are discussed and illustrated in Chapter 5, and quasi- and pre-experimental designs are examined in Chapter 6. Chapter 7 is concerned with measurement and quantification. The issues of validity and reliability—inseparable from measurement theory—
are also reviewed here, together with the issue of measurement error. In Chapter 8, we
present the principles of sampling theory, the most frequently used sampling designs,
and the methods for estimating sample size.
In Chapters 9 through 13, we present and illustrate the various methods of data
collection available to social scientists. A discussion of observational methods, including
laboratory experiments, field experiments, and natural experiments, as well as the importance of triangulation are the subjects of Chapter 9. Survey research—particularly the
mail questionnaire, the personal interview, the telephone interview, and the online survey—is examined in Chapter 10. Chapter 11 describes and illustrates methods of questionnaire construction: the content of questions, types of questions, question format, and
the sequence of questions. Chapter 12 is devoted to the theory and practice of qualitative
research, with a particular emphasis on participant observation and field research. In
Chapter 13, we discuss secondary data analysis—its utility and limitations, private and
public sources of secondary data, methods for searching, and content analysis.
The subsequent five chapters are concerned with data processing and analysis. In
Chapter 14, we present techniques of codebook construction, coding schemes and devices, ways to prepare data for analysis, the use of data-analysis software in social sciences research, and communication network linkages. Chapter 15 introduces the univariate distribution, measures of central tendency and dispersion, and various types
of frequency distributions. Chapter 16 examines bivariate analysis, focusing on the
relationship between two variables. Multivariate analysis is the subject of Chapter 17,
which presents ways of analyzing multiple variables with various statistical procedures.
Chapter 18 presents common techniques used in constructing indexes and scales; and in
Chapter 19, we discuss strategies of hypothesis testing, levels of significance, regions of
rejection, and parametric and nonparametric tests of significance.
This text, together with the supporting materials, will help readers move through the
major stages of the research process.
PREFACE
Student Online Resources
For the eighth edition, we have created a companion website to allow students
easier and more comprehensive access to resources and study aids. The online student resources provide the following study aids for each chapter: Chapter Abstract,
Learning Objectives, Key Terms, Flashcards, Web Quizzes, General Social Survey Data
Sets, and more.
We have moved a number of elements in the previous edition of the text to the companion website, such as the “Introduction to SPSS” (formerly Appendix A) now found
there. This introduction guides students through preparing and executing data analysis
using this widely available and often-used software package. The “Additional Reading”
sections at the end of each chapter in the previous edition are now located on the companion website as well. The “Sources for Research and Hypotheses” section at the end
of Chapter 3 in the previous edition has also been moved online. This section contains a
fully revised and updated listing of some of the most useful and current bibliographies,
indexes, journals, and statistical source books available for conducting social research.
Instructor Resources
The Instructor Resources to accompany Research Methods in The Social Sciences have been
expanded in the eighth edition revision. New resources include a Research Spotlight,
which provides current and provocative empirical exemplars to illustrate the themes
in each chapter and can be used for additional content, discussions, and student readings; new PowerPoint lecture slides for each chapter; and an expanded Test Bank. The
Instructor Resources also contain Essay and Discussion Questions, Research Projects,
and more.
Acknowledgments
Our literary debts are testified to throughout the text. Many students, instructors,
reviewers, and colleagues have offered useful ideas and comments since the first edition
was published.
We are particularly grateful to Michael Baer, Bruce S. Bowen, Jeffery Brudney, Gary
T. Henry, and Allen Rubin. We are also grateful to the Literary Executor of the late Sir
Ronald A. Fisher, F.R.S.; to Dr. Frank Yates, F.R.S.; and to the reviewers of the eighth
edition: Erika Austin, University of Alabama, Birmingham; Yoko Baba, San Jose State
University; Joseph Baker, East Tennessee State University; Edith Barrett, University of
Connecticut; Michelle Bemiller, Walsh University; Janet Bennion, Lyndon State College; Clancy Blair, New York University; James Bowie, Northern Arizona University;
Linda Dawson, University of Washington, Tacoma; Kathleen Dolan; University of Wisconsin, Milwaukee; Beth Donnellen, Kaplan University; Lee Dutter, Barry University;
Jonathan Eastwood, Washington & Lee University; Ted Fernandez, John Jay College;
Douglas Forbes; University of Wisconsin, Steven Point; Richard Fording, University of
Alabama; Bruce. J. Frayman, Aquinas College; Lidia Garrido, Barry University; Huston Gibson, Kansas State University; Elisabeth Gidengil, McGill University, Montreal;
Justina Grayman, New York University; Lewis Griffin, University of Denver; Rebecca
Gullman, Gwynedd-Mercy College; Sean-Shong Hwang, University of Alabama, Birmingham; Neal Jesse, Bowling Green State University; Chris Jones-Cage, College of the
Desert; Terry Keller, Lourdes University; Jo Anna Kelly, Walsh University; Phil Kelly,
Emporia State University; David Kriska, Walden University; Mona Lynch, University of
California, Irvine; Jeff Mullis, Emory University; Byron Orey, Jackson State University;
xiii
xiv
PREFACE
Philip Paolino, University of North Texas; Hyung Lae Park, Jackson State University;
Sejal Patel, Ryerson University; Jose Perez, Barry University; Gregory Petrow, University
of Nebraska, Omaha; Cindy Poag, Central Michigan University; C. Cybele Raver, New
York University; Malcolm Ree, Depaul University; Robert Rinkoff, Ryerson University;
Gisela Salas, Barry University; Juan Sandoval, St. Louis University; Ness Sandoval, St.
Louis University; David Schultz, Hamline University; Michael Smith, McGill University,
Montreal; Dale Story, University of Texas, Arlington; Stephen Sussman, Barry University; William Taggart, New Mexico State University; Melissa Tandel, Central Michigan
University; Joyce Tang, Queens College, City University of New York; John Taylor, Florida State University; George Taylor, Elon University; Clayton Thyne, University of Kentucky; Ogwo J. Umeh, California State University, East Bay; Francis Wayman, University of Michigan, Dearborn; Susan Wiley, George Washington University; and Rowena
Wilson, Norwalk State University.
We are grateful to Nina Reshef of The Public Policy Program, Tel-Aviv University,
who assisted David Nachmias, and whose significant contribution is evident throughout
the text; to Georgiann Davis, who assisted Chava Frankfort-Nachmias in the revision of
the text; and to Patrick Benner, who assisted Jack DeWaard in the revision and creation
of student and instructor resources.
Finally, we wish to express our indebtedness to the staff of Worth Publishers.
Chava Frankfort-Nachmias
David Nachmias
Jack DeWaard
RESEARCH METHODS
in the SOCIAL SCIENCES
this page left intentionally blank
PART I
FOUNDATIONS OF EMPIRICAL RESEARCH
The chapters in this book are grouped into four parts. Together, they illustrate the
sequential and often cyclical steps involved in social sciences research. In Part I, we
are concerned with the foundations of empirical research in the social sciences. What
are these foundations? The short answer is that there are many foundations. A more
involved answer is that the foundations of social research are not only diverse but also
highly integrated (that is, whole).
Social research cannot and should not be understood or undertaken without reference to early and vigorous debates in the philosophy of science, formal logic, and
consideration of different forms of logical reasoning (Chapter 1). Nor can social sciences
research occur without a common language in the form of clear and mutually agreed
upon concepts, definitions, theories, and models (Chapter 2). To make the connection
between the conceptual and observational worlds, social scientists must be equally rigorous in translating the above into a defined set of operations and procedures to specify
variables and hypotheses, measure and assess associations, and, ultimately, generate
knowledge (Chapter 3). With knowledge creation and dissemination also come important responsibilities with respect to the ethical conduct of research, including protecting
research subjects and conducting research and reporting results honestly and openly
(Chapter 4).
1
this page left intentionally blank
CHAPTER 1
THE SCIENTIFIC APPROACH
O WHAT IS SCIENCE?
O APPROACHES TO KNOWLEDGE
O BASIC ASSUMPTIONS OF SCIENCE
O AIMS OF THE SOCIAL SCIENCES
Scientific Explanation
Prediction
Understanding
O THE ROLES OF METHODOLOGY
Rules for Communication
Rules for Reasoning
Rules for Intersubjectivity
O SCIENTIFIC REVOLUTIONS
Normal Science
Revolutionary Science
A Logic of Discovery?
O THE RESEARCH PROCESS
O THE PLAN OF THIS BOOK
3
4
Part I
Foundation of Empirical Research
In 2011, the word “tweet” was formally incorporated into Merriam-Webster’s Collegiate Dictionary and defined as “(1) a chirping noise” and “(2) an online post made on
the social media site Twitter.” What can studies of tweets tell us about the nature and
structure of the social world? One recent paper attempted to examine several potential
insights.1 As part of its efforts to understand how actors other than mainstream media
outlets—including participants, ex-patriots, and other interested parties throughout the
world—participated in and portrayed the 2011 revolutions in Tunisia and Egypt, the
authors drew a detailed picture of the characteristics of these parties, their tweeting behaviors and contents, and their respective audiences (Twitter followers). In the process,
the study tracked nearly 1,000 tweeters and more than 400,000 tweets. The study’s
methodology was interdisciplinary by design and provided a blueprint for understanding the growing importance of social media as a serious avenue for reporting and communication that extends beyond the mainstream media. As these and other researchers
continue to analyze this exciting new data source and develop and assess different hypotheses, they will be following the rules of scientific methodology.
[
IN THIS CHAPTER we define science in relation to the search for knowledge and
proceed to discuss the basic assumptions of science and its aims and the roles of methodology. We define the scientific approach by its assumptions about nature and experience
and by its methodology as a set of tools for communication, reasoning, and intersubjectivity. We then discuss the ideas of scientific revolutions and discoveries. Last, we present
a model of the research process, the stages of which are discussed throughout this book.
We also ask a number of questions throughout this book: What benefits does the
scientific approach offer to people who take an interest in society? How can we acquire
knowledge about those aspects of the human experience that are social, political, economic, or psychological? For example, how can the scientific approach help us to understand phenomena ranging from inflation and unemployment to democratic governance
and bureaucracy, crime and delinquency, and even self-actualization? Our aim in this
chapter is to discuss why and how the social sciences are part of the family of sciences. ]
WHAT IS SCIENCE?
Science is a concept that is often misunderstood. Science is hard to define because we
tend to confuse the content of science with its methodology. While science has no specific
subject matter of its own, not every study of real-life phenomena is science. Astrologers,
for example, seek to establish relationships between significant life events and the positions of the stars in an effort to predict the future, but their goals, and the acts they perform in order to achieve them, do not qualify astrology for admission into the family of
sciences. Even if a university establishes a department of astrology and recruits faculty,
develops a curriculum, and offers a master of science degree, astrology would fail to
qualify as a scientific discipline. Why? Because the methodology used by astrologers is
regarded as unscientific. We therefore use the term science to refer to knowledge gathered by means of a distinctive methodology, namely a scientific methodology.
APPROACHES TO KNOWLEDGE
The word “science” is derived from the Latin word scire, which means “to know.” And
from the Greeks, the science of knowing is called epistemology. Philosophers have long
wrestled with ideas about how we can know what we claim to know. For example, how
Chapter 1
The Scientific Approach
does one truly know that the Sun will set each evening? That stepping on the accelerator
makes a car go faster? That smoking cigarettes causes cancer? These types of knowledge
claims are of a unique kind. When philosophers talk about knowledge, they are referring to propositional knowledge—descriptive claims that some state of affairs in the
world around us is either true or false. Thus, knowing how to ride a bicycle (referred to
as competence knowledge) or knowing your classmate’s name or favorite color (referred to
as acquaintance knowledge) is not the object of inquiry.
Like philosophers, scientists are squarely in the business of making propositional
claims to knowledge. In some cases, these knowledge claims (but not necessarily the
work involved in arriving at them) are very simple statements, such as that the Higgs
boson particle exists (versus does not exist). In other cases, knowledge claims concern
the truth value of associations and relationships, sequences and processes, and other dynamics in the social world. To take just one example, consider the recent and hotly contested propositional claim by Mark Regnerus that children whose parents have engaged
in same-sex relationships tend to be worse off on a number of outcomes compared with
children whose parents have not engaged in same-sex relationships:
Compared with children who grew up in . . . mother–father families, the children of women who reported a same-sex relationship [are] markedly different
on numerous outcomes, including many that are obviously suboptimal (such as
education, depression, employment status, or marijuana use).2
Aside from being highly controversial and potentially misleading,3 what makes
Regnerus’s claim propositional is the fact that it is a statement that some empirical phenomenon is true (or false).
Of course, not all propositional claims are equal. Some are better than others. How is
this so? It turns out that propositional knowledge is composed of three key ingredients:
O
Propositional knowledge is rooted in a set of beliefs.
O
These beliefs must be true.
O
These true beliefs must be justified.
In Regnerus’s case, the motivation for his paper was the idea, the belief, that same-sex
relationships and parenting might negatively impact children, and the desire to see if
this belief was consistent with reality, that is, whether this belief was true. He then went
to the data and conducted his analysis, the results of which provided him with empirical
evidence, or justification, for concluding that his belief that same-sex relationships and
parenting negatively impact children aligned with reality.
This is to say that we can think of scientists as producers of justified true beliefs—that
is, as developers and disseminators of knowledge. Of the three ingredients listed above,
justification has historically been the most controversial. In order to understand why,
imagine that you and a classmate are asked to predict whether it will rain tomorrow.
Based on the fact that it rained today and the day before, you predict that it will rain tomorrow. Your classmate, however, being more technologically savvy than you, pulls out
her smartphone and checks the latest weather report. After doing so, she also predicts
that it will rain tomorrow. Who has greater justification for their claim? Obviously, your
classmate does. She is justified on the basis of evidence from the weather report, while
your claim is merely based on a hunch that the future will resemble the past.
How does one obtain sufficient justification for a true belief and thereby acquire
knowledge? Early debates in epistemology concentrated on the roles of experience and
intuition. Empiricists believe that knowledge claims are justified on the basis of our
sensory experiences alone—sight, hearing, touch, and so forth. Knowledge of this sort
is called a posteriori (literally, “from the latter”) knowledge. Rationalists, on the other
5
6
Part I
Foundation of Empirical Research
hand, believe that knowledge claims are justified on the basis of rational intuition—
our innate capacity to grasp concepts and ideas regardless of our sensory experiences.
Knowledge of this variety is called a priori (or “from the former”) knowledge.
The German philosopher Immanuel Kant (1724–1804) argued for a hybrid of these
views.4 According to Kant, the world around us is chaotic. Our sensory perceptions
are what permit us to experience this chaos. Our rational faculties serve to organize
this chaos for us in meaningful ways. For example, scientists have repeatedly observed
that people with higher levels of education typically have higher levels of political involvement.5 According to Kant, while our sensory experiences might indicate a close
correspondence between education and political involvement, it is our rational intuition
that allows us to further classify this correspondence as one of cause and effect. Although
Kant’s work has not gone unchallenged, the utility of his approach was to provide an
account of knowledge that combined sensory experience and rational intuition in a way
that had not been attempted before.
Before we discuss the basic assumptions of science, it is worth noting that epistemological debates have expanded greatly since Kant. For example, the use of game theory in
political science embraces some of the core principles of rationalism. There are likewise
variants of game theory, such as evolutionary game theory, which incorporate empiricist ideas. Moreover, these epistemological debates and developments are not limited to
the issue of justification. For example, many scientists who use statistics in their research
adopt the view dating back to the English minister and mathematician Thomas Bayes
(1701–1761) that beliefs are not simply either/or propositions (e.g., either I believe it will
rain tomorrow or I do not), but instead are matters of degree and are updated as more
information is collected.6 This line of thinking underlies an important methodological
area of research in the social sciences, Bayesian statistics, which aims to incorporate uncertainty in our beliefs into statistical procedures and calculations.7
Likewise, regarding the truth component of knowledge claims, Karl R. Popper
(1902–1994) held that scientists must abandon attempts to provide evidence in favor
of competing claims to knowledge and instead focus their efforts on disproving (or falsifying) prevailing explanations.8 His idea was that knowledge claims that cannot be
falsified, and so stand the test of time, are the best candidates for truth. As with Bayesian
statistics, this line of thinking has also resulted in the development of specialized statistical procedures for working with data and testing hypotheses.9
BASIC ASSUMPTIONS OF SCIENCE
The scientific approach is grounded on a set of basic assumptions, fundamental premises considered to be unproven and unprovable. These assumptions are necessary prerequisites for conducting the scientific discourse.
1.
Nature is orderly. There is recognizable regularity and order in the world; events
do not just occur at random. Scientists assume that relationships and structures
continue to exist within rapidly changing environments. They likewise assume
that change is patterned and, therefore, can be understood. Order and regularity
are inherent in all phenomena. For example, there is no logically compelling
reason why the seasons should follow each other as they do, with winter following
autumn, autumn following summer, and so on. But, because winter always follows
autumn, despite the variations in temperature or snowfall, scientists conclude that
other regularities may likewise underlie other observable phenomena.
2.
We can know nature. This assumption expresses the conviction that human beings
are as much a part of nature as other phenomena. Although each of us possesses a
Chapter 1
The Scientific Approach
set of unique characteristics, as human beings we can be studied and understood
using the same methodology employed to study other natural phenomena. In
essence, individuals and social phenomena exhibit sufficient recurrent, orderly,
and empirically demonstrable patterns that are amenable to scientific investigation.
3.
All natural phenomena have natural causes. That all natural phenomena have
natural causes is at the core of the scientific revolution. By rejecting the belief
that forces other than those found in nature cause natural events, science
opposes fundamentalist religion, spiritualism, and magic. Scientists explain
the occurrence of phenomena in natural terms. As empirical regularities are
discovered and established, they can serve as evidence for the existence of
cause-and-effect relationships.
4.
Nothing is self-evident. Scientific knowledge is not self-evident. Accordingly,
claims for truth must be demonstrated objectively. Scientists cannot rely on
tradition, subjective beliefs, and cultural norms. They admit that possibilities
for error are always present; hence, even the simplest of knowledge claims call
for objective verification. Because of this characteristic, scientific thinking is
skeptical and critical.
5.
Knowledge is based on experience. If science is to help us understand the real
world, it must be empirical; that is, it must rely on our perceptions, experience,
and observations. Experience is an essential tool of the scientific approach, and
it is achieved through our senses:
Science assumes that a communication tie between man and the external
universe is maintained through his own sense impressions. Knowledge
is held to be a product of one’s experiences, as facets of the physical,
biological, and social world play upon the senses.10
However, it is also the case that many phenomena cannot be experienced or
observed directly. For example, on February 15, 2013, a meteor traveling at 11 miles
per second disintegrated over the Russian town of Chelyabinsk with the seismic
force of roughly 30 atomic bombs.11 In order to understand how and why this
event occurred, scientists developed ideas and sophisticated computer models of
how the orbits of the meteor and Earth came to collide, the angle and velocity with
which the meteor entered Earth’s atmosphere, and the accompanying shock wave
that was felt by people on the ground. In order to provide this account, scientists
also had to rely on rational intuition as a guide, grounded in scientific terms,
concepts, theories, and models. As Karl Popper once wrote:
The naive empiricist . . . thinks that we begin by collecting and arranging
our experiences, and so ascend the ladder of science. . . . But if I am ordered: “Record what you are experiencing,” I shall hardly know how to
obey this ambiguous o
rder. Am I to report that I am writing; that I hear
a bell ringing; a newsboy shouting; a loudspeaker droning; or am I to
report, perhaps, that these noises irritate me? . . . A science needs points
of view, and theoretical problems.12
6.
Knowledge is superior to ignorance. Closely related to the assumption that we can
know ourselves as well as we can know nature is the belief that knowledge should
be pursued for its own sake and for its contribution to improving the human
condition. The contention that knowledge is superior to ignorance does not,
however, imply that everything in nature can or will be known. Scientists assume
that all knowledge is tentative and changing. Things that we did not know in the
7
8
Part I
Foundation of Empirical Research
past we know now, and what we consider to be knowledge today may be modified
in the future. Truth in science is always dependent on the evidence, methods,
and theories employed, and it is always open to review. The belief that relative
knowledge is better than ignorance is diametrically opposed to the position taken
by approaches based on absolute truth. As Gideon Sjoberg and Roger Nett put it:
Certainly the ideal that human dignity is enhanced when man is restless,
inquiring, and “soul searching” conflicts with a variety of belief systems
that would strive toward a closed system, one based on absolute truth.
The history of modern science and its clash with absolute systems bears
testimony to this proposition.13
True believers already “know” all there is to know. In contrast, scientific knowledge
threatens the old ways of seeing and doing things; it challenges dogma, stability, and
the status quo. In return, the scientific approach can offer only tentative truth, whose
validity is relative to the existing state of knowledge. The strengths and weaknesses of
the scientific approach rest on the provisional and relative nature of truth:
It is a strength in the sense that rational man will in the long run act to correct
his own errors. It is a weakness in that scientists, not being so confident of the
validity of their own assertions as is the general public, may, in those frequent
periods when social crises threaten public security, be overturned by absolutists.
Science is often temporarily helpless when its bastions are stormed by overzealous proponents of absolute systems of belief.14
AIMS OF THE SOCIAL SCIENCES
Having discussed the assumptions of science, we are now in a position to address the
question raised earlier: What does science have to offer people who take an interest
in society’s problems? The ultimate goal of the social sciences and all other sciences is
to produce a cumulative body of verifiable knowledge. Such knowledge enables us to
explain, predict, and understand the empirical phenomena of interest to us. We believe
that a substantial body of knowledge can be used to improve the human condition. But
what are scientific explanations? When can we make predictions? When are we justified
in claiming that we understand empirical phenomena?
Scientific Explanation
Social scientists attempt to provide explanations for “why” questions—that is, why a
phenomenon has occurred and the set of conditions that caused it. The term explanation
thus refers to the process of relating a phenomenon to be explained to one or more other
phenomena. For example, why are government expenditures per capita so much higher
in Great Britain than in the United States? One response could be that the British want
their government to spend more. Such an explanation might satisfy the layperson, but
it would not satisfy social scientists unless the same reasoning explained the level of
government expenditures per capita in other countries. In fact, despite reports that most
Britons want their government to spend more, government expenditures per capita in
Great Britain declined after the Conservative Party returned to power in the 1980s.
Carl G. Hempel distinguished between two basic types of scientific explanations—
deductive and inductive.15 His classification is based on the kinds of generalizations
afforded by each type of explanation. Charles Sanders Peirce (1839–1914) proposed a
third type of scientific explanation—abductive—that is often used in the social sciences.16
Chapter 1
The Scientific Approach
DEDUCTIVE EXPLANATIONS. While fairly uncommon in the social sciences, in
deductive explanations a scientist explains an observed phenomenon by showing that
it must follow from an established universal law. For example, let us assume that it is a
universal law that all elected officials in a democracy will seek reelection. Suppose that Jane
Brown is an elected official in the United States in the year 2014. From this universal law,
we would conclude that Jane Brown will seek reelection. Moreover, we can arrive at this
conclusion from the comfort of our living room because deductive explanations do not
require going out into the world to make observations and analyze data.
In a deductive explanation, when the conclusion follows from the various parts or premises of the explanation as dictated by formal logic—that is, a system of reasoning in the development and evaluation of explanations—the explanation possesses deductive validity. When
the conclusion is likewise generated from premises that are true (versus false), the explanation
is also deductively sound. Returning to our example of Jane Brown, this explanation is valid
(indeed, in formal logic this type of explanation is called a deductive syllogism), but it is not
sound because it is simply false that all elected officials will seek reelection in a democracy.
Of course, deductive explanations are not always this rigid. Suppose that we revise
our universal law to say that some elected officials in a democracy will seek reelection,
or that it is possible that a person will seek reelection if she is an elected official in a
democracy. These sorts of contingencies are the objects of two domains of formal logic
called predicate and modal logic, respectively. For our purposes, however, the central takeaway message is that conclusions derived from deductive explanations are true by virtue of whether their premises are true. And their premises are considered true until it is
demonstrated that they are not.
INDUCTIVE EXPLANATIONS. Not all scientific explanations are (or can be) deductive.
This is particularly the case in the social sciences, where few, if any, universal laws have
been established. Social scientists often use inductive explanations (also known as
probabilistic explanations). Recalling our earlier example of Jane Brown, suppose that a
social scientist observes that elected officials in a democracy will seek reelection 75% of
the time. After all, elected officials are nothing more than human beings, who occasionally
do not seek reelection for a host of reasons due to family situations, retirement, and
scandal. Whereas there might be a strong correspondence between being an elected
official in a democracy and subsequently seeking reelection, this relationship cannot be
expressed as a universal law because not all elected officials will in fact seek reelection.
Inductive explanations are derived from probabilistic generalizations that express
an arithmetic relationship between two phenomena (e.g., n percent of X 5 Y) or the
tendency for two phenomena to take place simultaneously (e.g., X tends to correspond
to Y). When compared to universal laws in deductive explanations, the primary limitation of inductive, or probabilistic, explanations is that conclusions cannot be drawn with
complete certainty; thus, inductive explanations cannot be logically valid (or invalid).
In inductive explanations, when the conclusion is likely to logically follow from the
premises of the explanation, the explanation is said to be inductively strong (versus
weak). When a conclusion strongly follows from premises that are also true, the explanation is said to be inductively sound. Recalling our example of Jane Brown, an inductive
argument would go as follows: Research finds that elected officials in a democracy will
seek reelection about 75% of the time. Jane Brown is an elected official in the United
States in the year 2014. Thus, it is likely that Jane Brown will seek reelection. In our case,
the probability of her doing so is about 0.75, or 75%.
ABDUCTIVE EXPLANATIONS. Social scientists are often not in the position to make
inductive explanations. This is usually the case when scientists study what are referred
to as hard-to-reach populations such as victims of sex trafficking and undocumented
9
10
Part I
Foundation of Empirical Research
immigrants; other topics that are difficult to research because of hard-to-reach populations
include HIV prevalence among men who have sex with men and, in a recent study by
Alice Goffman, the struggles of young black men in urban Philadelphia, who are current
or former felons, in avoiding police apprehension and being returned to prison. In these
instances, crafting explanations is like fine detective work and involves constructing an
informed explanation about one or more causes from an observed social phenomenon.
Abductive explanations (also known as retroductive explanations) are employed in
situations when deductive explanations are impossible to make (because no universal
laws exist) and inductive explanations are not feasible (because there is not enough data to
support probabilistic generalizations). Abductive explanations also tend to be employed
when researchers want to provide a more “on the ground” empathetic understanding of
social phenomena. For example, in Goffman’s study, she observed the following:
Three weeks before Alex was due to complete his parole . . . a man with a
hooded sweatshirt covering his face stepped quickly out from behind the side
of a store and walked Alex, with a gun in his back, into the alley . . . took his
money and p
istol-whipped him three times, then grabbed the back of his head
and smashed his face into a concrete wall. . . . He refused [to go to the hospital],
saying that his parole officer might hear of it and serve him a violation for being
out past curfew.17
From this observed outcome—that is, Alex’s refusal to go to the hospital—Goffman
later concludes that modern parole and, more generally, criminal justice systems are
“organized not on the principle of occasional fear-inspiring public brutality, but on a
panoptic system of inspection, surveillance, and graded rewards and punishments.”18
In essence, Goffman’s explanation is as follows: “I observe that Alex refuses to go
to the hospital after this particular incident. If it is true that people under constant police surveillance refuse to do things like go to the hospital out of fear of sanction and/
or apprehension, then one reasonable explanation for Alex’s refusal to go to the hospital is that he believes (consciously or not) that he is constantly under surveillance.”
Goffman makes similar arguments with respect to Alex and other study participants in
their avoiding work, court appointments, and even contact with certain family members
and friends.
One of the risks associated with conclusions from abductive, or retroductive, explanations is that the observed phenomenon is, in fact, caused by something outside the
purview of the scientist. For example, perhaps Alex refused to go to the hospital because
(and did not report to Goffman that) he owed and had not yet repaid a $500 debt to one
of the nurses working at the hospital that evening. Considerations of this sort render
abductive explanations, which are reasoned from observed phenomena to one or more
possible causes, potentially risky, because the truth of the premises does not guarantee
that the conclusion will follow, even if they are the only option for studying certain hardto-reach populations.
Prediction
Deductive, inductive, and abductive explanations are components of scientific knowledge. Prediction is another. The ability to make correct predictions is a fundamental
characteristic of science. For example, if you know that the freezing point of water is
32°F, then you can predict what will happen if you fail to add antifreeze to the water in
your car radiator during the winter. Likewise, if you know that governments typically
increase spending during economic recessions, then you can predict that future recessions will likely bring about increases in government expenditures.
Chapter 1
The Scientific Approach
That scientific knowledge should lead to accurate predictions is based on the argument that if you know that X causes Y and that X is present, then you can predict that Y
will occur. Underlying this argument is the assumption that if the conditions required
for a phenomenon have been met, then the only reasons for failing to make an accurate
prediction are (1) that the statement X causes Y is not in fact true, or (2) that the antecedent (prior) condition, X, has not been correctly identified. For example, let us say
that chronic unemployment continues after establishing job placement programs. In this
case, either the idea that job placement programs solve unemployment problems is untrue, or the set of job placement programs that have been implemented are not really job
placement programs—for example, despite being labeled as job placement programs,
when implemented, they largely function as drop-in centers where people can browse
employment ads online without receiving any job placement training or counseling.
The logical structure of explanation is identical to that of prediction. The difference
between them lies in the perspective of the scientist. In the case of explanation, a scientist
seeks to relate an observed phenomenon in the past or present to a set of antecedent conditions to arrive at an explanation for what she has observed. In the case of prediction,
a scientist already knows the antecedent conditions and, on their basis, seeks to understand a future phenomenon.
Understanding
The third essential component of social scientific knowledge is understanding. The term
understanding is used in two very different ways—Verstehen (empathic understanding) and predictive understanding. These different usages evolved as the social sciences
progressed as sciences. Most important, the terms reflect the subject matter of the social
sciences—human behavior—and the fact that social scientists are more than just observers; they are participants in the world they study. In the words of Hans L. Zetterberg:
Symbols are the stuff out of which cultures and societies are made. . . . For
example, a sequence of conception, birth, nursing and weaning represents the
biological reality of parenthood. But in analyzing human parenthood we find, in
addition to the biological reality, a complex of symbols [e.g., values, norms] dealing with the license to have children, responsibilities for their care and schooling,
rights to make some decisions on their behalf, obligations to launch them by certain social rituals. . . . Our language thus contains codifications of what parents
are and what they shall do and what shall be done to them, and all these sentences in our language represent the social reality of parenthood. Social reality, in
this as in other cases, consists of symbols.19
But are symbols and, by implication, human behavior amenable to investigation by
the same methodology used in the natural sciences? Is the subject matter of the social
sciences so complex and distinctive that a unique methodology is required for its study?
Do social scientists, unlike natural scientists, have to “get inside” their subject matter in
order to understand it?
THE VERSTEHEN TRADITION. According to the Verstehen (a German term meaning
“empathy”) tradition, the natural and social sciences comprise separate bodies of
knowledge since the nature of their subjects is different. Proponents of the Verstehen
tradition believe that natural and social scientists should employ different research
methodologies. For example, unlike the natural scientist, a social scientist must grasp
both the historical dimension of social phenomena and the subjective aspects of the
human experience. The German sociologist Max Weber (1864–1920) argued that if
social scientists want to understand the behavior of individuals and groups, they must
11
12
Part I
Foundation of Empirical Research
“put themselves into the place of the subject of inquiry.”20 They must comprehend the
other’s view of reality, the way that reality is expressed in symbols, and the values and
attitudes that underlie these views.
More recently, the interpretive approach has emerged as an outgrowth of the Verstehen tradition. Kenneth J. Gergen, one of the major proponents of this approach, stated:
A fundamental difference exists between the bulk of the phenomena of concern
to the natural as opposed to the sociobehavioral scientist. There is ample reason
to believe that the phenomena of focal concern to the latter are far less stable
(enduring, reliable, or replicable) than those of interest to the former.21
Scientists who adopt the interpretive approach realize that not only is their subject
matter different, but, because of this difference, the credibility of the findings and
explanatory principles that they propose are considered to be less than those attributed
to the natural sciences. The methodology of Verstehen and its approaches are elaborated
in Chapter 12.
PREDICTIVE UNDERSTANDING. In contrast to the Verstehen tradition, many social
scientists hold the view that we can, in fact, attain objective knowledge when studying
the social world. Logical empiricists contend that the natural and social sciences can be
investigated using the same methodology. Although they acknowledge that empathic
understanding is useful in making scientific discoveries, they argue that these discoveries
should be validated by empirical observations before they are integrated into the larger
body of scientific knowledge. The idea of discovery versus validation is discussed in
more detail later in this chapter.
THE ROLES OF METHODOLOGY
The sciences are united not by their subject matter, but by methodology. A scientific
methodology is a system of rules and procedures that provides the foundations for
conducting research and evaluating claims to knowledge. This system is neither static
nor infallible. Rather, these rules and procedures are constantly reviewed and improved
as scientists look for new means of observation, analysis, inference, and generalization.
Once procedures are found to be compatible with the underlying assumptions of the
scientific approach, they are incorporated into this system of rules and, thus, the “logic
of inquiry” that governs scientific methodology. Hence, scientific methodology is first
and foremost self-correcting:
Science does not desire to obtain conviction for its propositions at any price. . . .
[A] proposition must be supported by logically acceptable evidence, which must
be weighed carefully and tested by the well-known canons of necessary and
probable inference. It follows that the method of science is more stable, and more
important to men of science, than any particular result achieved by its means.
In virtue of its method, the scientific enterprise is a self-corrective process. It
appeals to no special revelation or authority whose deliverances are indubitable
and final. It claims no infallibility, but relies upon the methods of developing and
testing hypotheses for assured conclusions. The canons of inquiry are themselves
discovered in the process of reflection, and may themselves become modified
in the course of study. The method makes possible the noting and correction of
errors by continued application of itself.22
The methodology of the social sciences continues to evolve slowly and carefully.
During its evolution, the exchange of ideas, information, and criticism makes it possible
to firmly establish, or institutionalize, commonly accepted rules and procedures, and to
Chapter 1
The Scientific Approach
develop corresponding methods and techniques. This system of rules and procedures
represents the normative framework of scientific methodology. Scientific norms set the
standards to be followed in scientific research and analysis. In other words, they define
the “rules of the scientific game.” These rules, in turn, enable communication, promote
constructive criticism, and enhance scientific progress.
Rules for Communication
Anatol Rapoport illustrated the general problem of communication that exists between
two people who have not shared a common experience by means of the following
anecdote:
A blind man asked someone to explain the meaning of “white.”
“White is a color,” he was told, “as, for example, white snow.”
“I understand,” said the blind man. “It is a cold and damp color.”
“No, it doesn’t have to be cold and damp. Forget about snow. Paper, for instance,
is white.”
“So it rustles?” asked the blind man.
“No, indeed, it need not rustle. It is like the fur of an albino rabbit.”
“A soft, fluffy color?” the blind man wanted to know.
“It need not be soft either. Porcelain is white, too.”
“Perhaps it is a brittle color, then,” said the blind man.23
Following the lines of this anecdote, we can say that a major purpose of methodology in the social sciences is to help scientists “see,” to facilitate communication between
researchers who either have shared or want to share a common experience. Methodology, however, does more. By making its rules explicit, public, and accessible, it creates
a framework for replication and constructive criticism. Replication—the repetition of
an investigation by the same or other scientists, in exactly the same way it was performed earlier—represents a safeguard against unintentional error and, sometimes,
deception.24 Constructive criticism implies that as soon as claims for knowledge are
made, we can ask the following questions: Does the explanation (or prediction) follow
logically from the assumptions? Are the observations accurate? What observational
methods were used? Was the testing procedure valid? Did any factor interfere when
drawing the conclusions? Should the findings be taken as evidence that another explanation is possible? Throughout this book we show that such questions embody criteria for
evaluating claims for scientific knowledge, whether old or new. Scientific methodology,
therefore, provides the foundations for assuring that these questions are understandable
to everyone i nvolved in the study of social phenomena.
Rules for Reasoning
Although empirical observations are fundamental to the scientific approach, they
do not speak for themselves. Empirical observations or facts must be ordered and
integrated into systematic, logical structures. The essential tool of the scientific
approach, along with empirical observation, is logic—the system of valid reasoning
that permits its users to draw reliable inferences on the basis of those factual observations. Logical procedures take the form of closely interdependent sets of propositions that support each other. By using logic as the foundation of scientific thinking,
scientific methodology promotes the internal consistency required of scientific claims
for knowledge.
Applying scientific methodology (or methodologies) therefore requires c ompetence
in logical reasoning and analysis. In the following chapters we discuss the elements
13
14
Part I
Foundation of Empirical Research
of logic—the rules for definition, classification, forming deductive as well as inductive
inferences. We also discuss the rules of scientific reasoning, theories of probability,
sampling procedures, systems of calculus, and measurement, the elements that constitute the social scientist’s methodological toolkit. It is important to remember throughout that the use of logic enables science to progress in a systematic and, sometimes,
revolutionary way.
Rules for Intersubjectivity
Scientific methodology provides a set of generally accepted criteria for determining
empirical objectivity (truth) and for choosing the appropriate methods and techniques
for validating evidence and conclusions. Objectivity and validation are interdependent. Empirical objectivity depends on validation, so much so that scientists cannot
make claims for objectivity until others have verified their findings. Intersubjectivity,
which is the sharing of observations and factual information among scientists
toward reaching consensus, is indispensable because logical reasoning alone cannot
guarantee objectivity.
As discussed in the previous section, logic is concerned with systematic reasoning
and does not necessarily deal with empirical truths or validated facts. Scientists can
make erroneous inferences from validated facts if they reason incorrectly. In deductive
and inductive explanations, these mistakes take the form of invalid and weak arguments, respectively. Scientists can also make erroneous inferences if they reason correctly (deductive validity or inductive strength) but fail to employ validated facts, thus
rendering their explanations unsound. As aptly noted by Anatol Rapoport, “The truth
of an assertion is related to experience; the validity of an assertion is related to its inner
consistency or its consistency with other assertions.”25
Truth cannot be established on purely logical grounds and must be confirmed by
empirical evidence. Thus, the term intersubjectivity more accurately describes this
process than does objectivity. In order to be intersubjective, knowledge must be communicable and mutually agreed upon by members of the scientific community. The significance of intersubjectivity lies in the ability of scientists to understand and evaluate
the methods used by others, and to conduct similar studies to replicate one another’s
work. If one scientist conducts a study, another scientist must have the methodological
toolkit at their disposal to replicate it and to compare the two sets of findings. If the
methodology is correct and the conditions under which the study was conducted are
stable, we expect the two sets of findings to be identical. Although conditions may
change and new circumstances emerge, as observed by Abraham Kaplan, “The methodological question is always limited to whether what is reported as an observation
can be used in subsequent inquiry even if the particular observer is no longer a part
of the context.”26
SCIENTIFIC REVOLUTIONS
The importance of scientific methodology is found primarily in what it provides—a toolkit and set of rules for observation and validation. Understood in this way, science is
prescriptive and demands conformity. Scientists reject claims for knowledge that do not
conform to the rules and procedures as prescribed by scientific methodology. But can
methodological conformity hinder us from making new discoveries and, by implication,
scientific progress?
Philosophers of science and social theorists have long been concerned with the
dangers of conformity and dogma in science. Among the various attempts made to
Chapter 1
The Scientific Approach
describe the progress of science from a sociological-political perspective, Thomas S.
Kuhn’s well-known and provocative theory of scientific communities is worth outlining
in some detail.
Normal Science
Kuhn termed normal science the routine verification of a theory (or paradigm) dominant in any historical period. Here, validation and testing are chiefly puzzle-solving
activities. In Kuhn’s words:
“Normal science” means research firmly based upon one or more past scientific achievements, achievements that some particular scientific community acknowledges for a time as supplying the foundation of its practice. Today such
achievements are recounted, though seldom in their original form, by science
textbooks, elementary and advanced. These textbooks expound the body of accepted theory, illustrate many or all of its successful applications, and compare
these applications with exemplary observations and experiments.27
Scientific textbooks socialize students and potential practitioners into the scientific
community. They define the practice of science, the kinds of problems to be investigated,
the assumptions and concepts to be employed, and the types of research methods to be
used. Historically, such textbooks and research did so by sharing “two essential characteristics. Their achievement was sufficiently unprecedented to attract an enduring group
of adherents away from competing modes of scientific activity. Simultaneously, it was
sufficiently open-ended to leave all sorts of problems for the redefined group of practitioners to resolve.”28
Kuhn uses the term paradigm for theories and models that share these two a ttributes.
He suggests that paradigms are closely related to the idea of normal science:
[A]ccepted examples of actual scientific practice—examples which include law,
theory, application, and instrumentation together—provide models from which
spring particular coherent traditions of scientific research. . . . The study of paradigms . . . is what mainly prepares the student for membership in the particular
scientific community with which he will later practice.29
Examples of paradigms might include Karl Marx’s dialectical materialism or Sigmund
Freud’s theory of the sexual origins of personality.
Furthermore, because scientists join a professional scientific community whose leaders have all learned the conceptual and methodological foundations of their discipline
from the same sources, their research will rarely evoke disagreement or criticism over
its fundamentals. Scientists whose research is grounded in a shared paradigm are psychologically committed to the same rules, norms, and standards of scientific practice:
“That commitment and the apparent consensus it produces are prerequisites for normal
science, [that is,] for the genesis and continuation of a particular r esearch tradition.”30
According to Kuhn, normal scientific communities are not neutral scientists; rather,
they are groups of partisans advocating and defending an established paradigm. While
Kuhn argued that scientific paradigms are often incommensurable, meaning that it is difficult to compare paradigms to one another given the absence of common units of measurement, adherence to a particular paradigm should not necessarily arrest scientific
progress. Paradigms, as organizing principles, are indispensable; without them, scientific research could not take place as a collective enterprise: “Acquisition of a paradigm
and of the more esoteric type of research it permits is a sign of maturity in the development of any given scientific field.”31 However, because normal science perpetuates itself,
it also has the potential to constrain change and innovation.
15
16
Part I
Foundation of Empirical Research
Revolutionary Science
In contrast to normal science, Kuhn views revolutionary science as the replacement
of one paradigm by another. Such change will take place gradually because it will
only gradually be accepted, and not without conflict, by the scientific community. For
example, the paradigm that human intelligence is a product of both the sociocultural
environment and genetic processes transformed the earlier paradigm of intelligence as
determined entirely by genetic processes. The new paradigm revolutionized the study
of personality and human behavior; it later became the cornerstone of many innovative
social, educational, and economic policies in the public sphere.
According to Kuhn, rejection of a dominant paradigm is a process that begins with
attempts to verify it. As scientists test empirically the diverse dimensions and implications of a dominant paradigm, its congruence with those research findings becomes tenuous. At some point a rival paradigm is constructed. Conflict then emerges between the
supporters of the old paradigm and the proponents of the new. Eventually, the scientific
community accepts the new paradigm and returns to those activities typical of normal
science. Such a transition may take decades.
Scientific revolutions are therefore rare. Scientists devote most of their time to normal science. As a rule, they do not try to refute dominant paradigms; nor do they immediately perceive their anomalies. Scientists, like other professionals, see what they expect
to see. For this reason, a dominant paradigm tends to remain the accepted paradigm
long after it fails to be congruent with empirical observations.
A Logic of Discovery?
In sharp contrast to Kuhn’s descriptive view of science and a scientific community that
is inherently more resistant to change is Karl Popper’s prescriptive, normative theory,
advocating for what science should be. Popper argues that a scientific community ought
to be and, to a considerable degree, actually is an “open society” in which no dominant
paradigm is ever sacred. Popper states that science should be in a state of permanent
revolution and that criticism lies at the heart of any scientific enterprise. For him, refutations of claims for knowledge in the form of attempts to falsify knowledge claims
constitute the essence of its revolutions:
In my view the “normal” scientist, as Kuhn describes him, . . . has been badly
taught. He has been taught in a dogmatic spirit: he is a victim of indoctrination. He has learned a technique which can be applied without asking for the
reason why.32
Popper does admit, however, that at any given moment, all scientists are “prisoners”
caught in their paradigms, expectations, past experiences, and language—with one important qualification:
We are prisoners in a Pickwickian sense: if we try, we can break out of our framework at any time. Admittedly, we shall find ourselves again in a framework,
but it will be a better and roomier one; and we can at any moment break out of
it again.33
In order to place these two points of view into perspective, we need to distinguish
etween two contexts of scientific activity, justification and discovery.34 The context of
b
justification refers to performing specific activities to validate knowledge claims both
logically and empirically. The scientific method is the basic tool used in this context
and provides scientists with a logic for doing so regardless of how they arrived at their
insights into the subjects they study. In contrast, the context of discovery represents
Chapter 1
The Scientific Approach
the process by which new insights are made, unconstrained at first by methodology.
Scientific methodologies may facilitate making discoveries but, at the initial stages of
exploration, no formal rules or logic can operate as perfect guides. This is the point
where creativity, insight, imagination, and inspiration are enormously important.
THE RESEARCH PROCESS
Scientific knowledge is knowledge grounded in both reason and experience (observation). These two criteria are translated into scientific research through the research
process. The research process is a set of activities in which scientists engage to generate
knowledge; it is the paradigm of scientific inquiry.
As illustrated in Figure 1.1, the research process consists of seven fundamental stages:
problem definition, hypothesis construction, research design, measurement, data collection, data
analysis, and generalization. Each stage influences the development of theory and is influenced by it in turn. In this book, we will discuss each stage, as well as the transitions between them. For the moment, we will limit ourselves to a general overview of the process.
The most characteristic feature of the research process is its cyclical nature. It typically begins with a problem and ends with a tentative empirical generalization. The
generalization ending one cycle serves as the beginning of the next cycle. This cyclical
process continues indefinitely and reflects the progress of a scientific discipline and the
growth of scientific knowledge.
Problem
Definition
Hypothesis
Construction
Generalization
Theory
Data
Analysis
Research
Design
Data
Collection
Measurement
FIGURE 1.1 THE MAIN STAGES OF THE RESEARCH PROCESS
17
18
Part I
Foundation of Empirical Research
The research process is also self-correcting over time. Scientists test tentative generalizations, called hypotheses, about research problems both logically and empirically.
If they reject these generalizations, they formulate and test new ones. In the process,
scientists reevaluate the research operations that they have performed because of errors
in how the research was conducted. For example, a researcher might reject the generalization that an economic crisis leads to increased government spending if it cannot be
logically or empirically verified. That is, a generalization can be rejected even if it is true if
the procedures for validation and verification (research design, measurement, and data
analysis) are deficient. To minimize the risk of rejecting true generalizations, scientists
reexamine each stage of the research process before revising old generalizations or suggesting new ones. This is why scientific methodology is said to be self-correcting. Ideas
and theories, by themselves, do not provide the tools for reexamining their basic claims
or the conclusions they are meant to support.
Finally; one should be aware that the research process as presented here is somewhat
idealized; that is, it is a highly general reconstruction of scientific practice:
[R]econstruction idealizes the logic of science only in showing us what it would
be if it were extracted and refined to utmost purity. . . . [But] not even the greatest
of scientists has a cognitive style which is wholly and perfectly logical, and the
most brilliant piece of research still betrays its all-too-human divagations.35
In practice, the cyclical nature of the research process occurs, as William L. Wallace states:
(1) sometimes quickly, sometimes slowly; (2) sometimes with a very high degree of
formalization and vigor, sometimes quite informally, unself-consciously, and intuitively; (3) sometimes through the interaction of several scientists in distinct roles
(of, say, “theorist,” “research director,” “interviewer,” “methodologist,” “sampling
expert,” “statistician,” etc.), sometimes through the efforts of a single scientist; and
(4) sometimes only in the scientist’s imagination, sometimes in actual fact.36
Therefore, our idealized reconstruction of the research process is not intended to be
rigid; it is meant to convey the underlying themes of social sciences research.
THE PLAN OF THIS BOOK
This book is organized to follow the major stages of the research process. Chapters 2 and
3 cover the conceptual foundations of empirical research and the relationships between
theory and research. They focus on the fundamental elements of research: concepts,
definitions, the functions and structures of theories, models, relations, variables, and the
construction of research hypotheses.
Chapter 4 is concerned with the ethical and moral questions confronted by social
scientists. In that chapter we discuss issues relating to the rights of research participants,
the obligations of scientists, the interactions between participants and scientists, and
professional codes of ethics. Those issues have become increasingly important in planning and executing research projects.
In Chapters 5 and 6, we focus on the research design stage. A research design is the
strategy that guides the investigator throughout the process of research. It is a logical
model of proof that allows the researcher to draw inferences concerning the causal relations that may be found among the phenomena under investigation. As you will see,
there are various types of research designs, each of which includes the conditions for
accepting or rejecting causal inferences.
Chapter 7 is concerned with the measurement stage of research. At this stage, researchers systematically assign symbols, usually numbers, to empirical observations.
Chapter 1
The Scientific Approach
19
These numbers and their relations are amenable to quantitative analyses capable of
revealing information that could not have been discerned without them. Numbers,
because they can be added, subtracted, percentaged, and correlated, are used for describing, explaining, and predicting social phenomena.
Typically, scientific generalizations are not based on all the measured observations
that might be obtained but on a relatively small number of cases—a sample. In Chapter 8,
we cover the major topics involved with sampling theory: methods for choosing representative samples, sample size, and sample designs.
The five subsequent chapters, Chapters 9 through 13, cover the data collection
stage. This is the stage where researchers make and record empirical observations. Data
(observations) can be collected by various methods, including structured observation,
nonstructured observation, personal interviews, impersonal surveys, public records, or
private records. No data collection method is foolproof, nor will any one method suit all
research problems. Different problems call for different methods, and each method has
inherent advantages as well as limitations.
Chapter 14 focuses on the major aspects of data processing, the link between data
collection and data analysis. During data processing, the observations researchers have
gathered in the data collection stage are transformed into a system of conceptual categories. These categories are then translated into coding schemes that also lend themselves
to quantitative analysis. These codes can then be recorded and processed by computers.
The central issues involved in coding and automatic data processing are also covered in
that chapter.
In the next stage of the research process, scientists conduct quantitative, statistical
analyses. Statistics are numbers that can be used to summarize, analyze, or evaluate
a body of information. It is useful to distinguish between two categories of statistics
according to their different functions: descriptive statistics and inferential statistics. Researchers use descriptive statistical procedures to organize, describe, and summarize
data. Chapter 15 covers descriptive univariate distributions; Chapter 16, bivariate distributions; and Chapter 17, multivariate data analysis techniques. In Chapter 18, we
present methods of index construction and scaling. The second category of statistics, inferential or inductive statistics, makes it possible for researchers to generalize beyond the
data in hand, to evaluate differences among groups, and to estimate unknown values.
These methods, discussed in Chapter 19, facilitate the conduct of systematic inquiry.
SU M M A RY
1. The sciences are united by their methodology,
not by their subject matter. What sets the
scientific approach apart from other ways
of acquiring knowledge is the assumptions
on which it is grounded in addition to its
methodology.
2. The assumptions of the scientific approach
are as follows: Nature is orderly, we can
know nature, natural phenomena have
natural causes, nothing is self-evident,
knowledge is derived from the acquisition
of experience, and although knowledge is
tentative, it is superior to ignorance.
3. The methodology of the scientific approach
serves three major purposes: It provides rules
for communication, rules for logical and valid
reasoning, and rules for intersubjectivity.
This system of rules allows us to explain,
predict, and understand human behavior
and events in our environments.
4. The scientific method requires strict
adherence to the rules of logic and
observation. Such adherence discourages
dogma because it maintains that the
research process is cyclical and selfcorrecting. Rational criticism should be at
20
Part I
Foundations of Empirical Research
the heart of the scientific enterprise, and
science ought to be in permanent change.
Claims for knowledge are ultimately
accepted only insofar as they are congruent
with the assumptions and methodology of
science.
K E Y T E R MS FO R R E VI E W
abductive explanations (p. 10)
a posteriori (p. 5)
a priori (p. 6)
assumptions (p. 6)
constructive criticism (p. 13)
context of discovery (p. 16)
context of justification (p. 16)
deductive explanations (p. 9)
deductive validity (p. 9)
deductively sound (p. 9)
empirical (p. 7)
empiricists (p. 5)
epistemology (p. 4)
explanation (p. 8)
inductive explanations (p. 9)
inductively sound (p. 9)
inductively strong (p. 9)
interpretive approach (p. 12)
intersubjectivity (p. 14)
justification (p. 5)
knowledge (p. 5)
logic (p. 9)
logical empiricists (p. 12)
normal science (p. 15)
paradigm (p. 15)
prediction (p. 10)
propositional knowledge (p. 5)
rationalists (p. 5)
replication (p. 13)
research process (p. 17)
revolutionary science (p. 16)
science (p. 4)
scientific methodology (p. 12)
Verstehen (p. 11)
STU D Y Q UE S TI O N S
1. What is knowledge? How do science
and scientific methodology help to create
knowledge?
2. Discuss the basic assumptions underlying
the scientific approach.
3. What are the aims of science as a knowledgeproducing enterprise?
4. Compare and contrast the three logics—
deductive, inductive, and abductive—
discussed in this chapter. What are the
hallmarks of each? Which of these logics is
most consistent with the Verstehen tradition?
5. What does it mean to say that scientific
methodology provides a set of rules,
e.g., for communication, reasoning, and
intersubjectivity?
6. Describe the research process and its stages.
7. How is science actually carried out, (a) as a
cyclical process of reasoning and observation,
and (b) institutionally?
RE A D I N G A N D W R I TI N G R E S E A R C H REPORT S
1. On your own or in small groups, see if you can
brainstorm two to three general research topics
that are of interest to you, such as violence at
sporting events, immigration, the impact of
technology on political participation, and so
on. Identify what you find most interesting
about these topics. Did you select these
topics purely as a scholarly exercise? Did any
important personal experiences lead you to
select the topics that you did?
2. For each of the two to three topics that you
selected, see if you can come up with several
C h a p t e r 1 T h e S c i e n t i f i c A p p r o a c h
propositions that you might later investigate
in an empirical research project. For
example, if you expressed an interest in the
topic of immigration to the United States,
perhaps you are interested in the specific
idea that immigration is bad for the U.S.
economy, for example, because immigrants
take away jobs from native-born workers,
use up social welfare benefits, and so on.
you—such as bibliographies, indexes,
abstracts, peer-reviewed journals, etc.—
conduct a brief but focused search on the
topics you identified in the previous steps.
See if you can find any examples of studies
that address one or more of the topics you
identified. What sorts of propositions did
these studies consider? File these studies
away so that you can access them at a later
time. You might also take a look at the
references provided in these studies to see if
any of these are of interest or useful to you.
3. Now it is time to consult the research
literature. Using the Web, your school
library, and other resources available to
O
21
ONLINE STUDY RESOURCES
Visit the Research Methods in the Social Sciences, 8th edition, website at
http://bcs.worthpublishers.com/rmss8e for additional chapter-specific
study aids, including practice quizzes, learning objectives, flash cards,
General Social Survey data sets, additional readings, suggested research
sources, and step-by-step instructions to using SPSS.
R E F E R E N C ES
1.
Gilad Lotan, Erhardt Graeff, Mike Ananny,
Devin Gaffney, Ian Pearce, and Danah
Boyd, “The Arab Spring! The Revolutions
Were
Tweeted:
Information
Flows
During the 2011 Tunisian and Egyptian
Revolutions,” International Journal of
Communications 5 (2011): 1375–1405.
2.
Mark Regnerus, “How Different Are the
Adult Children of Parents Who Have
Same-Sex Relationships? Findings from
the New Family Structures Study,” Social
Science Research 41, no. 4 (2012): 764.
3.
The Regnerus paper prompted a letter
to the editor of the journal Social Science
Research on behalf of 200 concerned
scholars: Gary J. Gates et al., “Letter to
the Editors and Advisory Editors of Social
Science Research,” Social Science Research
41, no. 6 (2012): 1350–1351. Among their
concerns, these scholars pointed out
several serious problems with respect
to how Regnerus defined same-sex
relationships and parents, as well as
inconsistencies in the peer-review process
of the journal Social Science Research.
4.
Immanuel Kant, Critique of Pure Reason,
trans. Max Muller (London: Macmillan,
1881), 688.
5.
Cindy D. Kam and Carl L. Palmer,
“Reconsidering the Effects of Education on
Political Participation,” Journal of Politics
70, no. 3 (2008): 612–631. See also John
Henderson and Sara Chatfield, “Who
Matches? Propensity Scores and Bias in the
Causal Effects of Education on Participation,”
Journal of Politics 73, no. 3 (2011): 646–658.
6.
Thomas Bayes, “An Essay Towards Solving
a Problem in the Doctrine of Chances,”
Philosophical Transactions 53 (1763): 370–418.
7.
Bruce Western, “Bayesian Analysis for
Sociologists: An Introduction,” Sociological
Methods and Research 28 (1999): 7–34.
8.
Karl R. Popper, The Logic of Scientific
Discovery (New York: Routledge, 1992).
22
9.
Part I
Foundations of Empirical Research
Kent D. Miller and Eric W. K. Tsang,
“Testing Management Theories: Critical
Realist Philosophy and Research Methods,”
Strategic Management Journal 32, no. 2 (2011):
139–158.
10. Gideon Sjoberg and Roger Nett, A
Methodology for Social Research (New York:
Harper & Row, 1968), 26.
11. Paul Rincon, “Russia Meteor’s Origin
Tracked Down,” BBC News, February 26,
2013. See also Rebecca Jacobson and Ellen
Rolfes, “Meteor Explodes Over Central
Russia Triggering Destructive Sonic
Blast,” PBS Newshour, February 15, 2013.
12. Popper, The Logic of Scientific Discovery.
13. Sjoberg and Nett, A Methodology for Social
Research, 25.
14. Ibid., 26.
15. Carl G. Hempel, Philosophy of Natural
Science (Englewood Cliffs, NJ: PrenticeHall, 1966), Chapter 5.
16. Charles S. Peirce, “Illustrations of the
Logic of Science,” Sixth paper, “Deduction,
Induction, and Hypothesis,” Popular
Science Monthly 13 (1878): 470–482.
17. Alice Goffman, “On the Run: Wanted
Men in a Philadelphia Ghetto,” American
Sociological Review 74, no. 3 (2009): 345.
18. Ibid., 354.
19. Hans L. Zetterberg, On Theory and
Verification in Sociology, 3rd enl.. ed.
(Totowa, NJ: Bedminster Press, 1965),
1–2. See also Kenneth J. Gergen, Toward
Transformation in Social Knowledge (New
York: Springer-Verlag, 1982).
20. Max Weber, The Theory of Social and Economic
Organization, trans. A. M. Henderson and
Talcott Parsons (New York: Free Press, 1964).
21. Kenneth J. Gergen, Toward Transformation
in Social Knowledge (New York: SpringerVerlag, 1982), 12.
22. Morris R. Cohen and Ernest Nagel, An
Introduction to Logic and Scientific Method
(New York: Harcourt Brace & World,
1962), 395–396.
23. Anatol Rapoport, Operational Philosophy:
Integrating Knowledge and Action (San
Francisco: International Society for General
Semantics, 1969), 12.
24. See for example the recent case of Marc
Hauser who was found to have falsified data
related to the recognition patterns of monkeys
in his studies. Office of Research Integrity,
“Findings of Research Misconduct,” Notice
Number: NOT-OD-12-149 (Department
of Health and Human Services, National
Institutes of Health, 2012).
25. Rapoport, Operational Philosophy, 18.
26. Abraham Kaplan, The Conduct of Inquiry:
Methodology of Behavioral Science (New
Brunswick, NJ: Transactions, 1998), 128.
27. Thomas S. Kuhn, The Structure of Scientific
Revolutions, 2nd ed. (Chicago: University
of Chicago Press), 10.
28. Ibid.
29. Ibid.
30. Ibid., 10–11.
31. Ibid., 11.
32. Karl R. Popper, “Normal Science and
Its Dangers,” in Criticism and the Growth
of Knowledge, eds. Imre Lakatos and
Alan Musgrave (New York: Cambridge
University Press, 1970), 53.
33. Ibid., 56.
34. Kaplan, The Conduct of Inquiry, 12–18.
35. Ibid., 128.
36. Walter L. Wallace, The Logic of Science in
Sociology (Chicago: Aldine), 19.
CHAPTER 2
CONCEPTUAL FOUNDATIONS OF RESEARCH
O CONCEPTS
Functions of Concepts
O DEFINITIONS
Conceptual Definitions
Operational Definitions
Example: The Definition of Alienation
Bridging the Conceptual and Operational Divide
O THEORY: FUNCTIONS AND TYPES
What Theory Is Not
Four Types of Theories
Axiomatic Theory
O MODELS
Example: A Model of Policy Implementation
Example: A Model of the Dynamics of Racial Systems
O THEORY, MODELS, AND EMPIRICAL RESEARCH
Theory Before Research
Research Before Theory
23
24
Part I
Foundations of Empirical Research
What is ethnic identity? This is the central question motivating a recent book by Kanchan
Chandra, professor of politics at New York University, on constructivist theories of ethnic
politics. In day-to-day usage, we often talk about and treat ethnic groups as if they were
real, discrete, and substantive entities in the world. For example, one can’t help thinking
about the importance of the Hispanic vote in President Barack Obama’s reelection in 2012.
But, according to Chandra, conceptualizing ethnic groups in this way is highly problematic, because it is based on the assumption that “ethnic identities that describe individuals
and populations are singular, timeless and fixed for all time.”1 Accordingly, Chandra argues that the concept of ethnic identity must be understood as a hybrid of two distinct but
related concepts. First, Chandra defines ethnic structure to refer to what she calls “nominal
descent-based attributes” such as skin color or hair texture. She then distinguishes ethnic
structure from the concept of ethnic practice, which refers to how descent-based attributes
are “activated” in daily life, for example, in electoral politics. This conceptualization implies that ethnic identity is both a property of individuals and a process that is bounded by
and contingent on the contexts in which individuals and groups find themselves.
The latter observation invites discussion of the historical conditions surrounding how
ethnic identities become (and also cease to become) salient, and has served as the basis for
related efforts by Chandra and others to measure ethnic identity in empirical research.2
[
IN THIS CHAPTER we first discuss concepts, the building blocks of theoretical
s ystems. After reviewing the functions of concepts, we discuss conceptual and operational definitions, and how they relate to the research process. We then distinguish
among the four levels of theory and explain how models represent different aspects of
the real world. We close this chapter by reviewing the debate over theory-then-research
and research-then-theory as the appropriate strategy to adopt when doing research in
the social sciences.
As we saw in Chapter 1, scientific knowledge is validated by both reason and experience. This implies that social scientists operate at two distinct but interrelated levels—the
conceptual (theoretical) and the empirical (observational). Social sciences research is the
outcome of the interaction between these two levels. In this chapter we focus on the conceptual level and the relationships that exist between its components and empirical research. ]
CONCEPTS
Thinking involves the use of language. Language itself is a system of communication composed of symbols and the rules that permit us to combine these symbols in different ways.
One of the most significant symbols in a language, especially as it relates to research, is the
concept. A concept, like other symbols, is an abstraction, a representation of an object, or
one of that object’s properties, or a behavioral phenomenon. Scientists begin the process
of research by forming concepts as shorthand descriptions of the empirical world. Each
scientific discipline develops its own unique set of concepts. For example, “social status,”
“role,” “power,” “bureaucracy,” “community,” “relative deprivation,” and “cohort” are
common concepts in political science and sociology, whereas “intelligence,” “perception,”
and “learning” are common in psychology. To scientists, concepts and symbols constitute
a professional language. Thus, when a social scientist uses the word “cohort,” other social
scientists immediately know what it represents: a group of people sharing a demographic
characteristic such as age. For people who are untrained in the social sciences, “cohort”
might be gibberish, because the term has no meaning for them.
Chapter 2
Conceptual Foundations of Research
Functions of Concepts
Concepts serve many important functions in social sciences research. Among these
functions, first and foremost, concepts are tools for communication. Without a set of
agreed-upon concepts, scientists could not communicate their findings or replicate
one another’s studies. As we saw in Chapter 1, communication in the social sciences
is achieved via intersubjective sharing and agreement. Without concepts, this would
simply be impossible.
Second, concepts afford researchers a perspective. Concepts are a lens through which
to view and subsequently make sense of the world around them: “Through scientific
conceptualization the perceptual world is given an order and coherence that could not
be perceived before conceptualization.”3 Concepts enable scientists to focus on some
aspect of reality by defining its components and then by attempting to discover whether
that aspect is shared by different phenomena in the real world:
It permits the scientist, in a community of other scientists, to lift his own idiosyncratic experiences to the level of consensual meaning (i.e., intersubjectivity). It
also enables him to carry on an interaction with his environment; he indicates to
himself what a concept means and acts toward the designation of that meaning.
The concept thus acts as a sensitizer of experience and perception, opening new
realms of observation, closing others.4
Third, concepts permit both classification and generalization. Scientists use concepts to
categorize, order, and generalize from their observations of empirical phenomena. As
John C. McKinney puts it:
To introduce order with its various scientific implications, including prediction, the scientist necessarily ignores the unique, the extraneous, and
[the] nonrecurring, and thereby departs from perceptual experiences. This
departure is the necessary price he must pay for the achievement of abstract
generality. To conceptualize means to generalize to some degree. To generalize means to r educe the number of objects by conceiving of some of them as
being identical.5
Finally, concepts are components of theory. When concepts are linked in logical and
systematic ways, they lead to theories. Concepts are the building blocks of any theory
because they ultimately define its content and attributes. The concepts of “power” and
“legitimacy,” for example, are central to theories of governance, while the concepts
of “individualism” and “Protestantism” lie at the core of Émile Durkheim’s theory of
suicide. These examples illustrate the very close correspondence between concept formation and theory construction.
Four Functions of Concepts
O
Concepts provide a common language, which enables scientists to communicate
with one another.
O
Concepts give scientists a perspective—a way to look at phenomena and make
sense of the world around them.
O
Concepts allow scientists to classify and order their observations and experiences
and to generalize from them.
O
Concepts are components of theories—they define a theory’s content and
attributes.
25
26
Part I
Foundations of Empirical Research
DEFINITIONS
If concepts are to be useful, they have to be clear, precise, and agreed upon. Language,
however, is often vague or ambiguous. Concepts such as “power,” “bureaucracy,”
“discrimination,” and “satisfaction” mean different things to different people and are
used in different contexts to designate various things. Usually, this does not create major
problems in ordinary conversations. Science, however, cannot progress with an ambiguous and imprecise language.
Because of this need for precision, every scientific discipline is concerned with its vocabulary. To help them in achieving the desired clarity and precision during research, scientists
employ two major types of definitions: conceptual and operational. The difference between
them is important, for they help the scientist decide how variables are to be measured.
Conceptual Definitions
Definitions that describe concepts by using other concepts are called conceptual definitions. In certain cases, these definitions may be used as variables in the research process.
For example, “power” has been conceptually defined as the ability of an actor (e.g., an
individual, a group, the state) to get another actor to do something that the latter would not
otherwise do. The conceptual definition of “relative deprivation” is an actor’s perception of
a discrepancy between his or her “value expectations” and his or her “value capabilities.”6
In the latter example, “value expectations” and “value capabilities” are themselves
concepts: We cannot sense value expectations empirically; we can only conceive of them
through intellectual processes of abstraction. However, the process of definition need not
stop here. In the case of relative deprivation, a person who is unfamiliar with the term is
likely to ask, “What are values, capabilities, expectations, and perceptions?” These concepts call for further clarification. Expectations, for example, have been defined as manifestations of the prevailing norms found in the immediate economic, s ocial, cultural,
and political environments. But what is meant by “norms,” “immediate,” “economic,”
“social,” “cultural,” and “political”? These are also concepts that need to be defined by
still other concepts, and so on.
At some point in this progression, concepts need not be defined using other concepts.
In such cases, the concept is said to be defined by primitive terms. Primitive terms are
unambiguous and are often conveyed using clear-cut empirical examples. For instance,
with respect to the example of power, the concept of “value capabilities” is sometimes
defined using primitive terms that express the level of material resources available to the
individual, such as disposable income.7,8
When primitive terms are combined and used to define a concept, the concept is said
to be defined using derived terms. Derived terms are those that are defined using primitive
terms in an integrated way. For example, if there is agreement on the meanings of such primitive terms as “individual,” “interact,” and “regularly,” then one can subsequently define
the concept of “group” (a derived term) as two or more individuals who interact regularly.
As is evident, derived terms make it possible to refer to increasingly complex phenomena in
a highly efficient way. After all, when studying two or more individuals who interact regularly, it is more efficient to simply call this arrangement a group than to continually refer to
this arrangement using the three primary terms of individual, interact, and regularly.9
Last, it is important to realize that conceptual definitions are neither true nor false.
Concepts acquire meaning and must be assessed within the context of the theory in
which they are found. Conceptual definitions must be evaluated with respect to their
contribution to the theories in which they are embedded. What ultimately renders one
conceptual definition “better” than another depends on the theoretical context and the
clarity, coherence, and applicability of the definition therein.
Chapter 2
Conceptual Foundations of Research
Operational Definitions
As shorthand mental representations, concepts do not actually exist in the real world.
Concepts are abstractions; they are symbols of empirical phenomena that a researcher
intends to study. In the social sciences, treating concepts as though they were the phenomena that they represent is called the fallacy of reification. This is the error of regarding abstractions as real, rather than as a product of human thought.10
Operational definitions provide one such tool to bridge the conceptual (theoretical) and empirical (...
Purchase answer to see full
attachment